The immune reaction to colorectal cancer (CRC) strongly affects its recurrence and mortality. Cancers with high densities of CD8 cytotoxic, Th1, T-regulatory and memory T-cells have a better prognosis than cancers with a Th17, Th2 response. Whether immune factors have a similar impact on early, pre-invasive, carcinogenesis has not been studied. There are 3 common CRC precursor lesions: tubular adenomas (TAs), tubulovillous/villous adenomas (TVs), and sessile serrated polyps (SSPs). These lesions are fairly well characterized molecularly, but little is known regarding the immune reaction to them, even though these occur at a time when carcinogenesis might be more easily countered than after invasive cancer has developed. Preliminary data suggest that as adenomas progress toward invasive cancer, the Th2 and Th17 immune responses increase, and the Th1, CD8+ cytotoxic response diminishes. Unfortunately, these data are derived from relatively small studies, with non-specific methodology. In addition, no study has considered how the immune contexture in any CRC precursor effects the risk of colorectal neoplasia on follow-up, and essentially nothing is known about the immune contexture in SSPs. Moreover, there is a paucity of data regarding the influence of personal and lifestyle factors on the immune responses. To clarify these issues, we propose to utilize the biospecimens and extensive data from three completed studies of colorectal neoplasia, in order to characterize the local immune responses in 500 TAs, 150 TVs, and 150 SSPs, and study associations with future neoplasia. We propose 3 specific aims. 1) To contrast TAs, TVs, and SSPs with regard to infiltration by major T-cell types and immune gene expression 2) To assess the associations of the immune response with precursor lesion progression (size), and 3) To study associations of the immune response in TAs with risk of new neoplastic lesions at follow-up, and evaluate whether any observed association is independent of CRC/precursor risk factors. In exploratory analyses we will also investigate the associations of the immune responses in CRC precursors with risk and protective factors (e.g. obesity or smoking and NSAID use or exercise, respectively). We hypothesize that, as in CRC itself, the Th17 and Th2 responses will have a pro-carcinogenic effect and so will be 1) greater in TVs than in TAs and SSPs, 2) directly associated with polyp size, and 3) independently predictive of risk of recurrent neoplasia. Conversely, we hypothesize that the strength of the Th1, CD8, T-regulatory and memory T cell responses will have the opposite associations. In summary, we will provide the first rigorous and detailed assessment of the immune response to colorectal precursors by studying variability in immune response by precursor type, assessing its impact on lesion growth and recurrence, and identifying epidemiologic correlates of the immune response. This analysis will provide insight into the mechanisms underlying early colorectal carcinogenesis, help identify patients at risk for recurrence, and suggest possible preventive interventions. ! .!!